P R E P A R A T ION A N D
P R 0 P E R T I E S 0F
d
(A - A -T)n
d (A
- T - T)"
Sarma, R., Silverton, E. W., Davies, D. R., and Terry, W. D. (1971), J. Biol. Chem. 246,3753. Schechter, B., Conway-Jacobs, A,, and Sela, M. (1971), Eur. J. Biochem. 20,321. Schechter, B., Schechter, I., and Sela, M. (1970), J. Biol. Chem. 245,1438. Schechter, J. (1971), Ann. N . Y. Acad. Sei. 190, 394. Schechter, J., and Berger, A. (1966), Biochemistry 5,3362. Schechter, J., Schechter, B., and Sela, M. (1966), Biochim. Biophys. Acta 127,438. Steiner, L. A., a n d b w e y , S. (1966),J. Biol. Chem. 241,231. Tumerman, L. A., Nezlin, R. S., and Zagyansky, Y.A. (1972), FEBS (Fed.Eur. Biochem. Soc.)Lett. 19,290. Warner, C., and Schumaker, V. (1970a), Biochem. Biophys. Res. Commun. 38,125. Warner, C., and Schumaker, V. (1970b), Biochemistry 9, 451. Werner, T. C., Bunting, J. R., and Cathou, R. E. (1972), Proc. Nat. Acad. Sci. U. S . 69,795. Yguerabide, J., Epstein, H. F., and Stryer, L. (1970), J. Mol. Biol. 51,573. Zipper, P. (1969), Acta Phys. Austr. 30, 143. Zipper, P. (1972), ActaPhys. Austr. 36,27.
I , 351. Marrack, J. R., and Richards, C. B. (1971), Immmology 20, 1019. Metzger, H. (1970), Annu. Rev. Biochem. 39,889. Nezlin, R. S., Zagyansky, Y. A., and Tumerman, L. A. (1970), J. Mol. Biol. 50, 569. Ohta, Y., Gill, T. J. 111, and Leung, C. S. (1970), Biochemistry 9,2708. Pecht, I., Givol, D., and Sela, M. (1972), J. Mol. Biol. 68,251. Pilz, I. (1970), Allg. Prakt. Chem. 21,21. Pilz, I. (1973), in Physical Principles and Techniques of Protein Chemistry, Leach, S. J., Ed., New York, N. Y., Academic Press, in press. Pilz, I., Kratky, O., and Karush, F. (1973), manuscript in preparation. Pilz, I., Puchwein, G., Kratky, O., Herbst, M., Haager, O., Gall, W. E., and Edelman, G. M. (1970)) Biochemistry 9, 211. Porod, G. (1951). KolloidZ. 124,83. Robbins, J. B., Haimovich, J., and Sela, M. (1967), Immunochemistry 4, 11. Rotman, M. B., and Celada, F. (1968), Proc. Nat. Acad. Sci. U . S. 60,660.
Preparation and Properties of the Repeating Sequence Polymer d( A-A-T),.d( A-T-T), t Robert L. Ratliff,* D. Lloyd Williams, F. Newton Hayes, E. L. Martinez, Jr., and David A. Smith
ABSTRACT: The polymer d(A-A-T),. d(A-T-T), has been prepared through a combination of organic chemical procedures and enzymatic synthesis with DNA polymerase from Escherichia coli. The physicochemical properties of the polymer have been compared with those of dA,. dT, and d(A-T), .d(A-T),. Nearest neighbor analysis agrees with the assigned structure. Absorbance U S . temperature measurements as well as analyt-
T
he synthesis of polydeoxyribonucleotides has provided models for study of DNA and its biological functions. Singlestranded polymers with homo, multiblock, and random hetero sequences have been prepared by use of the terminal transferase from calf thymus (Bollum et al.. 1964; Ratliff et al., 1967, 1968; Ratliff and Hayes, 1967; Kat0 et al., 1967). Another procedure for obtaining single-stranded polydeoxyribonucleotides is separation of the complementary strands of double-stranded polymers (Wells and Blair, 1967). Doublestranded structures have been obtained by de novo synthesis, from primed reactions, and by the use of template. The de noljo reactions using the DNA polymerase from Escherichia coli or Micrococcus luteus have yielded d(A-T),. d(A-T), (Schachman et al., 1960), dA,.dT, (Burd and Wells, 1970),
t From the Cellular and Molecular Radiobiology Group, Los Alamos Scientific Laboratory, University of California, Los Alamos, New Mexico 87544. Receioed August 22, 1973. This work was performed under the auspices of the U.S . Atomic Energy Commission.
ical buoyant density determinations in cesium chloride and cesium sulfate showed that d(A-A-T), d(A-T-T), has properties lying in between those of dA,. dT, and d(A-T),. d(AT),. Both strands of the d(A-A-T),. d(A-T-T), polymer were found to be transcribed by the RNA polymerase from E . coli to give r(A-A-U),.r(A-U-U),.
dC,.dG, (Radding et a[., 1962), dC,.dI, (Inman and Baldwin, 1964), and d(C-I),.d(C-I), (Grant et al., 1968). Primed amplification reactions have used complementary oligo- or polydeoxyribonucleotides that range from homo to repeating tetramer sequences (Wells et al., 1965, 1967a,b, 1970; Morgan, 1970). Seven of the possible repeating trinucleotide complementary double-stranded polydeoxyribonucleotides have been reported (Wells et al., 1967a; Morgan, 1970). The five remaining are d(A-A-T),. d(A-T-T-),,l d(C-C-G),. d(C-G-G),, d(A-
1
The abbreviated notation for complementary double-stranded
polydeoxyribotrinucleotides is in one of the two general forms allowed by the IUPAC-IUB Combined Commission on Biochemical Nomenclature (1970) and further incorporates an alphabetization rule that allows only one way to write the structure of each polymer. The left-
hand member of the notation is stated as the alphabetically earliest three-letter set of the six possibilities in the polymer. The right-hand member is then fixed to be complementary in reverse order of letters. B I O C H E M I S T R Y , VOL.
12,
NO.
24, 1 9 7 3
5005
RATLIFF
TABLE I : RF Values of Deoxyribonucleotides and Derivatives.
Solvent Compound
I
I1
0.128 2.6' 3.8' 5.0' 4.0'
I11
IV
v
0.195
2.8" 2.0" 0. 3" 1.2" 0 .24"
0,88'
0.099 0.49 0.41 1 .95b 2.8b 1.5b 1. 4 b 0.31 0 . 7 0 b 0.57b 0.28b
0.57 1.8b 1,5b 0 15 3.0'
0.27b 1.3'? 0 19 0.11 0.54, 0.56
e t a/.
propanol ( 2 :5 ) ; (11) 0.5 M NH40Ac (pH 3.9)-ethanol (2 :5 ) ; (111) 1 M NH40Ac (pH 7.5)-ethanol (2:5); (IV)l-propanolconcentrated NHDH-water (55 :10 :35); (V) 2-propanolconcentrated NH1OH-water (7 : 1 :2); and (VI) isobutyric acid- concentrated NHIOH-H~O(66 :1 :33). Chromatographic mobilities of various nucleotides and protected derivatives are given in Table I. Paper electrophoretic separation of nucleotides was performed on strips of Whatman No. 3MM paper using a Savant high-voltage, tank-type apparatus with isoparaffinic coolant and formate buffer (pH 3.8). Size distribution of synthetic polynucleotides was obtained by chromatography through Bio-Gel A-50m (8 cm2 X 82.4 cm) in an LKB Model 4901A column, equilibrated with 50 mhi triethylammonium bicarbonate (pH 8.5). Analytical buoyant density centrifugation was carried out in a Spinco Model E ultracentrifuge equipped with ultraviolet optics and a photoelectric scanner with a multiplexer attachment. Double-sector, 12-mm Kel-F centerpieces and the four-hole An-F Ti rotor were used. Ultracentrifuge runs were carried out at 44,000 r'pm and 25". All samples run in cesium chloride included a marker of E. coli DNA with a density of 1.703 &'ml (Vinograd et d.,1963; Wells and Blair, 1967). All runs in cesium sulfate included a marker of T-4 DNA with density of 1.443 g,ml (Erikson and Szybalski, 1963). Density values were calculated by the equation
Relative to d-PA. Relative to d-pT. where r,,> and r5 are the distances in centimeters from the axis of rotation to the center of the marker and sample bands, respectively. For cesium chloride runs a value of 1.19 x log was used for $ (Ift er d., 1961); for cesium sulfate runs 3 was 5.934 X lo8 (Chamberlin and Berg, 1964). Density values were calculated as the average of several scans made in each run. Enzjmes. Micrococcal DNase was purchased from Mann Experimental Procedures Research Laboratories. Alkaline phosphatase was obtained from Worthington Biochemical Corp., and its solution in Chemicals and Reagents. There were obtained from 0.05 M glycylglycine buffer (pH 8.0) was heated at boiling water SchwarzIMann deoxythymidine 5'-phosphate, deoxyadenosine 5 '-phosphate, deoxyadenosine 5 '-triphosphate, [ CC]- bath tempel~aturefor 10 min to destroy endonuclease activity. Spleen phosphodiesterase was prepared by the methods of deoxythymidine 5'-triphosphate, ['Cladenosine 5'-triphosHilmoe (1960) and Richardson and Kornberg (1964). Eschephate, and ['CC]uridine 5'-triphosphate; from P-L Biochemiirchiu coli DNA polymerase was prepared according to the cals, Inc., adenosine 5 '-triphosphate and uridine 5 '-triphosprocedure of Richardson et it/. (1964), and both the DEAEphate, from International Chemical and Nuclear Corp. [a-3*P]deoxyadenosine 5'4riphosphate and [a- 32P]deo~y- cellulose fraction and the phosphocellulose fraction were used The preparations had specific activities of 300-1 300 units/mg thymidine 5'-triphosphate ; from New England Nuclear Corp. [~t-~~P]adenonsine 5'4riphosphate and [~x-~~P]uridineof protein when assayed using d(A-T),L.d(A-T),, as template. RNA polymerase from E. coli was purified by a modification 5'-triphosphate; and from Aldrich Chemical Co., Inc., 2 mesitylenesulfonyl chloride, 2,4,6-triisopropylbenzenesulfonyl of the procedure of Chamberlin and Berg (1962), as described previously (RatlitT et ai., 1964), and had a specific activity of chloride, and dicyclohexylcarbodiimide. Each of the two 3000 units;mg of protein. sulfonyl chloride reagents was recrystallized from petroleum Preparurion of 1/ztennediare.s. 2-Cyanoethyl 5'-thymidylate ether before every use. was prepared by the method of Schaller and Khorana (1963) Hydracrylonitrile was obtained from Eastman Kodak Co. and isolated in quantitative yield as the calcium salt. 03'and was fractionally distilled under vacuum. Pyridine used as Acetylthymidine 5'-phosphate was prepared according to solvent in the chemical syntheses was distilled successively Khorana and Vizsolyi (1961). N-Benzoyldeoxyadenosine 5'from chlorosulfonic acid and solid KOH and was then dried phosphate was prepared by the procedure of Ralph and over a molecular sieve (Linde, type 3A). Cesium chloride Khorana (1961). 03'-Acetyl-N-benzoyldeoxyadenosine 5 'and cesium sulfate were obtained from the Harshaw Chemical phosphate was prepared by the procedure of Weimann et al. Co. Agarose for column chromatography (Bio-Rad Labora(1963). 2-Cyanoethyl N-benzoyl-5'-deoxyadenylate was pretories) was obtained from Calbiochem. pared by a procedure similar to that of Ohtsuka et al. (1965); Analytical Procedures. Paper chromatography was done on N-benzoyldeoxyadenosine 5'-phosphate (2.3 mmol) was Whatman No. 40 paper by the descending technique using the allowed to react with hydracrylonitrile (230 mmol) and difollowing solvent systems: (I) 1 M NH40Ac (pH 6.0)-2-
C-C),. d(G-G-T),, d(A-C-G),. d(C-G-T),, and d(A-G-C), . d(G-C-T)n. We present here the synthesis and characterization of d(A-A-T),. d(A-T-T), and a comparison of its physicochemical properties with the previously reported dA,. dT, and d(A-T)n.d(A-T)
5006
B I O C H E M I S T R Y , VOL.
12,
NO.
24, 1 9 7 3
PRE PAR ATION
A
N D P R 0P E R T I E S 0F d
(A - A
- T), d (A -T-T),
cyclohexylcarbodiimide (23 mmol) in 15 ml of anhydrous pyridine for 48 hr. After the addition of 25 ml of water, the reaction mixture was kept at 2" for 2 days. Dicyclohexylurea was removed by filtration, and the filtrate wab concentrated under vacuum with additions of pyridine. Chromatography in solvent I showed a major spot at R F 0.51 (5.1 relative to d-pA at Rr 0.10) and a trace of material at RF 0.42 (d-p-BzA reference, R F 0.25). T o the residual solution of d-CNEt-pBzA in hydracrylonitrile were added 0.5 M CaBr, in 95% ethanol (2.15 mmol) and 200 ml of acetone which resulted in partial precipitation of the calcium salt. With stirring, 400 ml of ether was added, and the mixture was cooled at 2' for 16 hr. The solid calcium salt was collected in a fritted funnel and washed with dry acetone several times (total volume 125 ml). A solution of the solid in dry methanol (100 ml) was added with agitation to 1500 ml of ether. After 12 hr at 2", the precipitated calcium salt was collected in a medium fritted funnel and dried under vacuum; weight 1.2682 g. Ultraviolet analysis (eB0 18.3 x loa) indicated 2 295 mmol of Ca d-CNEt-p-BzA; the acetone wash liquor (125 ml) contained 0.06 mmol. Chemica/ Synthesis and Polyiuerization of Deoxytrinrtcleotides. d(A-A-T) SERIFS, Dinucleotide d(CNEt-p-BzAp-BzA). An anhydrous pyridine (8 ml) solution of pyridinium d-CNEtp-BzA (2.0 mmol), pyridinium d-p-BzA (2.64 mniol), and 2.73 g (13.2 mmol) of dicyclohexylcarbodiimide was shaken with 1 g of pyridinium Dowex 50W-X4 resin for 4 5 days at room temperature. With cooling an equal volume of water was added, and the solution was refrigerated for 16 hr. Dicyclohexylurea was removed by filtration and washed repeatedly with 30% pyridine. The combincd filtrate, after extraction of dicyclohexylcarbodiimide with petroleum ether, was concentrated to 75 ml, cooled in ice, and treated with cold 2 N NaOH (75 ml) for 20 min. The cold solution was poured onto excess pyridinium Dowex 50W-X4 resin (240 mequiv), with stirring. The resin was removed by filtration and thoroughly washed with 5% pyridine (200 ml). The total filtrate was reduced under vacuum to 200 ml, and the pH was adjusted to 7.0 with 0.1 N triethylamine. The desired dinucleotide was isolated in two parts by anion exchange chron~atographyon a DEAE-cellulose column (6.5 X 50 cm, bicarbonate form) with a linear gradient of triethylammonium bicarbonate (ph 7.5), 0.005-0.3 M in 12 1. The yield of pure d(p-BzAp-BzA) was 19,300 A280 unita (€BO18 x l o 3 baaed on phosphorus analysis). Paper chromatography in solvent I1 showed a major species with R P 0.65 and a trace of material at RF 0.5. After hydrolysis with concentrated N H 4 0 H , the resulting d(pApA) was homogeneous in two solvents ( R , 0.47 in solvent IV and R P 0.52 in solvent VI). An anhydrous pyridine (15 ml) solution of d(p-BzApBzA) (0.882 mmol), hydracrylonitrile (88.2 mmol), and dicyclohexylcarbodiimide (8.82 mmol) was bhaken at room temperature for 54 hr. With cooling, water (35 ml) was added; after 16 hr at 2", dicyclohexylurea was removed by filtration and dicyclohexylcarbodiimide extracted with ether. The aqueous pyridine filtrate was evaporated with additions of pyridine, and the calcium salt of d(CNEt-p-BzAp-BzA) was prepared as described above for d-CNEt-p-BzA. The isolated product (0.86 mmol of dimer based on phosphorus analysis) was homogeneous in solvent 11, R F 0.69. Trinucleotide d(p-BzAp-BzApT). An anhydrous pyridine (20 d)solution of trihexylammonium d(CNEt-p-BzAp-BzA) (0.63 m o l ) , pyridinium d-pT(Ac) (3.78 mmol), and triisopropylbenzenesulfonyl chloride (9.45 mmol) was shaken for 6 hr at room temperature. The reaction mixture was cooled
in ice, water was added (20 ml), the pH was adjusted to 7.2 with 1 N NH40H, and the solution was stored at 2" for 15 hr. To eflect removal of the cyanoethyl and 03'-acetyl groups, the aqueous pyridine solution (75 ml) was cooled in ice, and cold 2 N NaOH (75 ml) was added. After 20 min at 0", the solution was poured with stirring onto an excess of pyridinium Dowex 5OW-4X resin. The resin was collected and washed thoroughly with 25 % pyridine. The total filtrate was reduced in volume under vacuum, the pH was adjusted to 7.0, and the mixture was chromatographed on a DEAE-cellulose column (5.2 x 60 cm, bicarbonate) at 4". Elution was with tr linear gradient of triethylammonium bicarbonate (pH 7.5), 0.0050.4 M in 16 1. The yield of the desired trinucleotide was 3800 A?79units (0.09 mmol). Paper chromatography in solvent I11 gave a major species at RF 0.199 with a trace of material at RF 0.092. After removal of the N-benzoyl groups with 9 N NH40H at 60" during 2 hr, only one spot was apparent at RF 0.030 (solvent 111). Following renioval of the 5'-phosphate group with alkaline phosphatase, the d(ApApT) (RF 0.51, solvent IV) was degraded with spleen phosphodiesterase. The ratio of d-Ap l o d-Thd was 2.06 :1.00. Polytrimer d(A-A-T) 1 5 . Polymerization of the protected trinucleotide was done according to the procedure of Narang et a/. (1967) except that the 25 % of 03'-protectedtrinucleotide was not included. The solution of trihexykammonium d(pBzAp-BzApT) (3180 A2,, units. 0.076 mmol) in dry pyridine was made anhydrous by repeated evaporations of dry pyridine ( 5 x 1 ml). The residue was dissolved in dry pyridine (1 ml), mesitylenesulfonyl chloride (0.684 mmol) was added, and the volume of solution was reduced by half. After 5 hr at room temperature, 25% pyridine (1 ml) and triethylamine (1 ml) were added. After 24 hr at 2", the benzoyl groups were removed with concentrated NHaOH (39 hr), and the solution was evaporated under vacuum to a small volume (5 ml, pH 8.4). The mixture was then chromatographed on a DEAEcellulose column (3.8 X 77 cm, bicarbonate) with a linear gradient of triethylammonium bicarbonate (pH 7.5) from 0.005 to 0.4 M in 12 1. After emergence of the probable dodecamer peak, the column was washed with 1 M triethylammonium bicarbonate to remove together all the longer oligomers. d(A-T-T) SERIFS. The following synthesis of the trinucleotide d(pTpTp-BzA) was prepared essentially according to the procedure of Ohtsuka and Khorana (1967) Dinucleotide d(CNEt-pTpT). To an anhydrous pyridine (6 ml) solution of the pyridinium salts of d(CNEt-pT) (1.5 nimol) and d-pT(Ac) (1.5 mmol) was added 7.5 mmol of dicyclohexylcarbodiimide. After 4 days at room temperature water (10 nil) was added, and the solution was left for 16 hr. The aqueous pyridine solution was diluted to 25 ml with water, filtered to remove dicyclohexylurea, and extracted with petroleum ether to remove dicyclohexylcarbodiimide. The aqueous and heated at 60' for 3 hr. phase was made 9 N in ",OH It was then cooled, concentrated to 20 ml under vacuum, and filtered to remove additional dicyclohexylurea. The filtrate was applied to a DEAE-cellulose column (6.5 x 46 cm, bicarbonate); elution was with a linear gradient of triethylamnionium bicarbonate (pH 7.5), 0.005-0.6 M in 16 1. The yield of d(pTpT) was 13,200 AZslunits (0.713.mmol). Paper chromatography in solvent 111 indicated a small contaminant at RF 0.11 in the product, Rrq0.21. To eliminate the contaminant, the product was rechromatographed on the same DEAE-cellulose column. The purified product amounted to 12,200 A?,,?units and was homogeneous in solvent I. An anhydrous pyridine (7 ml) solution of pyridinium d(pTpT) (0.66 mmol), hydracrylonitrile (66 mmol), and dicycloBIOCHEMISTRY , L . 12, VO
NO.
24, 1 9 7 3
5007
RATLIFF
hexylcarbodiimide (7.26 mmo ) containing dry pyridinium Dowex 50W-X4 resin (0.5 g) was kept at room temperature for 3 days. The reaction mixture was diluted to 40 ml with water and kept at room temperature for 24 hr; it was then filtered to remove dicyclohexylurea and resin and extracted with petroleum ether to remove dicyclohexylcarbodiimide. Paper chromatography a t this stage, solvent I, indicated a major product at R F 1.2 relative t o d(pTpT) and a minor product a t RF 3.6 relative to d(pTpT). The high RF material was proven be an ester of the type bis(CNEt)-d(pTpT) (Narang et a/., 1967) by complete deblocking with 9 N ",OH. This resulted in a single product at RF 0.22 in solvent 111, corresponding to d(p'TpT). Triethylamine equivalent to the phosphate anion was added to the solution which was then concentrated to a syrup under vacuum, The residue was dried by evaporation of pyridine (2 x 10 ml), dissolved in dry pyridine (10 ml), and added with stirring to 400 ml of ether and left at 2" for 15 hr. After centrifugation, the mother liquor was decanted from the precipitate which was resuspended in 400 ml of ether and collected and freed of ether under vacuum. The dry ester was dissolved in 5 0 x ethanol (30 ml) and assayed, amounting to 11,700 ,4267 units (0.632 mmol). Trinucleotide d(pTpTp-BzA). An anhydrous pyridine (6 ml) solution of trihexylammonium d(CNEt-pTpT) (0.31s mmol), pyridinium d-p-BzA(Ac) (1.5 mmol), and triisopropylbenzenesulfonyl chloride (3.75 mmol) was shaken for 8 hr. With cooling water was added, the p H was adjusted to 7.3 with 1 N N H 4 0 H ,and the solution was filtered. To the filtrate (52 ml) cooled in ice was added cold 2 N NaOH (52 ml). After 20 min at O", the solution was poured onto excess pyridinium Dowex 50W-X4 resin (130 ml) with stirring. The resin was collected and washed with 5x pyridine. The total filtrate was applied to a DEAE-cellulose column (4.4 x 75 cm. bicarbonate) and chromatographed at 4" with a linear gradient of triethylammonium bicarbonate (pH 7.5), 0.005-0.4 M in 16 1. The yield of the desired trinucleotide was 2590 /I2!, units (0.075 mmol), with the following spectral properties at p H 8.3:,,,A, 274 nm, A m i n 240 nm. After removal of the benzoyl group d(pTpTpA), R F 0.061 in solvent 111, was with 9 N ",OH, dephosphorylated with alkaline phosphatase. The resulting d(TpTpA), Rp. 0.212 in solvent V, was degraded with spleen phosphodiesterase. The observed ratio of d-Tp io d-Ado was 1.93 :1.00. Polytrimer d(A-T-T) 2 j. Polymerization of the protected trinucleotide was according to Narang et al. (1967), with the exception noted above. To an anhydrous pyridine (1 ml) solution of trihexylammonium d(pTpTp-bzA) (0.075 mmol) was added mesitylenesulfonyl chloride (0.8 mmol), and the volume was immediately reduced by half. After 5 hr at room tempera( 5 ml) were added, ture 25 % pyridine (1 ml) and 9 N ",OH and the mixture was left for 7 days. After evaporation of NHlOH under vacuum, the reaction mixture was chromatographed on a DEAE-cellulose column (3.8 x 77 cm. bicarbonate) with a linear gradient of triethylammonium bicarbonate (pH 7.5), 0.005-0.4 M in 12 I. After emergence of the probable dodecamer peak, the column was washed with 1 M triethylammonium bicarbonate to remove together all longer oligomers. Enzwiatic Syntliesis of d(A-A-T),. d(A-T-'I'),. The synthesis of d(A-A-T), . d(A-T-T)n using the oligonucleotide fractions d(A-A-T) z s and d(A-T-T) 2 as template was initially accomplished with the DEAE-cellulose fraction of the DNA polymerase from E. coli. The reaction mixture, in 2 ml of total volume, contained 50 pmol of potassium phosphate buffer (pH 7.5), 10 bmol of MgCh, 2 pmol of 2-mercapto-
5008
B I O C H E M I S T R YV O , L. 12,
NO.
24, 1 9 7 3
et
UI.
ethanol, 5 pmol each of dATP and dTTP, 50 mono-nmol' of each of the oligodeoxyribonucleotides as templates, 80 units of E . coli DNA polymerase, and [14C]dTTP as the labeled substrate. The reaction temperature was 15", and the course of polymer synthesis was examined by radioisotope incorporation into acid-insoluble material (Hayes et a/.. 1966). The utilization of triphosphates amounted to 0.7 in 24 hr, 4 in 48 hr, and 34% in 72 hr, at which time reaction was stopped by addition of 0.1 vol of 1 ", sodium dodecyl sulfate, and protein was extracted with phenol as described previously (Smith el u / . , 1970). After chromatography through a Bio-Gel A-50m column. the isolated d(A-A-T), d(A-T-T),, (2.32 monopmol, 23-fold net synthesis) was used as template in a largescale reaction. The composition of the reaction mixture (total volume 40 ml) was the same as above except that I50 units of the phosphocellulose fraction of the DNA polymerase from E. coli and 50 mono-nmol of the d(A-A-T),, d(A-T-T), were used per milliliter of the reaction mixture. The isolated yield of polymer was 45.8 mono-pmol. The synthesis kinetics of each of the three polymers containing deoxyadenosine and deoxythymidine were measured by incorporation of [ 3'?P]deoxythyniidylicacid from cy-labeled triphosphate into acid-insoluble material. To follow the synthesis of d(A-A-T),,' d(A-T-T),,, a 1-ml aliquot was removed from the large-scale reaction mixture described above, and [ G Y - ~ T ] ~ Twas T Padded. The rates of formation of dA,;dT. and d(A-T),).d(4-T), were also determined in I-ml reaction mixtures. Nearest neighbor analysis was done for cach of the polymers according to the procedures of Josse et al. (1961) and Rat18 et al. (1968). Size distributions were determined as described by Hayes and Mitchell (1969) and by Jang and Bart1 (1971). The absorbance-temperature transitions were determined using the procedure and apparatus previously reported by Hayes et ul. (1970). The method for following the kinetics of transcription of each of the deoxyribopolymers by RNA polymerase from E. coli was essentially as described by Morgan (1970). Results The initial synthesis of d(A-A-T),, . d(A-T-T), using the complementary chemically synthesized oligodeoxyribonucleotides as template and the DEAE-cellulose fraction of DNA polymerase from E. coli was done at 15". All attempts to obtain synthesis at 37" resulted either in no reaction or in formation of the alternating polymer d(A-T)n.d(A-T),. The long incubation period is attributed to both the low enzyme concentration (1.2 mono-nmol of template t o 1 unit of DNA polymerase) and the temperature (15'). The highest yields of primed synthesis using d(A-A-T), . d(A-T-T), were obtained with the phosphocellulose fraction of the DNA polymerase from E . coli, and this enzyme fraction was used for large-scale preparation of d(A-A-T),. d(A-T-T), as well as for kinetic studies and preparation of material for nearest neighbor analysis of all three polymers. The synthesis of d(A-T),.d(A-T), reached a maximum of 36 % utilization of deoxyribonucleoside triphosphates present in 48 hr, as measured by incorporation of [ C ~ - ~ ~ P ] into ~TTP acid-insoluble material (Figure 1). The extent of synthesis of d(A-A-T),,.d(A-T-T), was 437, in 48 hr, while that of dA,.
2 The quantities of oligomers or polymers are referred to in terms of moles of mononucleotide residues as determined by phosphorus analyses.
P RE P A R A T I O N A N D
TABLE 11:
P R 0 P E R T I E S 0F
d(A-
A
- T),, d (A - T - T)"
Nearest Neighbor Analysis of Each of the Three Isomeric Polymers.
a- 32P-LabeledRadioact. in Deoxyribonucleoside 3 '-Phosphatea
Template d A,. dT, d(A-T), * d(A-T), d(A-A-T), .d(A-T-T),
Labeled Triphosphate dATP dTTP dATP dTTP dATP dTTP
3 '-Deoxyadenylate cpm Z 42,500 99 5 500 0 3 100 0 2 147,200 99 7 12,800 32 6 85,300 65 5
3 '-Deoxythymidylate cpm % 200 0 5 143,800 99 7 45,800 99 8 500 0 3 26,500 67 4 44,900 34 5
a- 3ZP-LabeledRadioact. in Ribonucleoside 2'(3')-Phosphateb Labeled Triphosphate
rATP rUTP
2'(3')Adenylate cpm
2 '( 3 ')Uridylate cpm
x
-
15,600 33 8 43,100 66 4
~
x
30,500 66 2 21,800 33 6
a For analysis of d(A-A-T),.d(A-T-T),, two 1-ml aliquots were removed from the large reaction mixture to one of which was added [ c Y - ~ ~ P I ~ and A T P[ C Y - ~ ~ P ] ~to T Tthe P other. For the other two polymers, the reaction mixtures and temperatures were the same as described for kinetic reactions (Figure 1) except that two different preparations were performed: one with [a-3*P]dATP and the second with [ C X - ~ ~ P I ~ TThe T P .32P-labeleddA,.dT, reactions were stopped after 24 hr, while those of d(A-T),. d(A-T), and d(A-A-T),.d(A-T-T), were stopped after 72 hr. The reactions were terminated by placing them in a boiling water bath for 10 min. The samples were cooled and centrifuged, and the supernatants were dialyzed against 0.01 M sodium pyrophosphate and then distilled water to remove unreacted deoxyribonucleoside triphosphates. The polymers were degraded to their 3 'mononucleotides using micrococcal DNase and spleen phosphodiesterase and were separated by electrophoresis on paper (Josse et al., 1961; Ratliff et al., 1968). The radioactive spots were located with a Packard radiochromatogram scanner, cut out, and counted in a Packard Tri-Carb liquid scintillation counter. All counts per minute (cpm) have been rounded off to the nearest 100. Two 0.2-ml aliquots were removed from a 5-ml transcription reaction containing 5 mono-pmol of d(A-T-T),.d(A-T-T),, 5 Mmol each of rATP and rUTP, 200 pmol of Tris-HC1 buffer (pH KO), 20 pmol of magnesium chloride, 5 pmol of manganese chloride, 60 pmol of 2-mercaptoethanol, and 2000 units of RNA polymerase. To one aliquot was added [ c ~ - ~ ~ P ] ~and A TtoP the other [a-32P]rUTP.Incubation was at 37" for 4 hr. The 32P-labeledreactions were terminated by placing them in a boiling water bath for 10 min. The samples were cooled and centrifuged, and the supernatants were dialyzed against 0.01 M sodium pyrophosphate and then distilled water. The retentates were made 0.3 N in KOH by adding 0.06 ml of 5 N KOH and adjusting the total volume to 1.0 ml with water. Incubation at 37" for 18 hr was followed by neutralization through addition of 0.06 ml of 5 N HC1 to each solution. The 2 '(3')-ribonucleotides were separated by paper electrophoresis according to the procedure of Markham and Smith (1952). The total results are averages of results from four separate runs.
dT, reached a maximum of 34% in 24 hr followed by a rapid drop to 1 2 z in 36 hr. The nearest neighbor analyses for each of the three isomeric polymers containing exclusively deoxyadenosine and deoxythymidine are given in Table 11. The results show that, in polymers prepared using [ C ~ - ~ ~ P ] ~all A Tradioactivity P, was found in d-Ap from dA,.dT,, in d-Tp from d(A-T),.d(A-T),, and two-thirds in d-Tp and one-third in d-Ap from d(A-A-T), . d(AT-T),. When [ c ~ - ~ ~ P ] ~was T Tthe P labeled deoxyribonucleoside triphosphate, 100% of the label was found in d-Tp from dA,. dT,, 100% in d-Ap from d(A-T),.d(A-T),, and two-thirds in d-Ap and one-third in d-Tp from d(A-A-T),.d(A-T-T),. The nearest neighbor results show that the dinucleotide sequences in d(A-A-T),. d(A-T-T),, were one-third each d(ApT) and d(TpA) and one-sixth each d(ApA) and d(TpT), as expected. When ribonucleoside [~y-~~P]triphosphates were used with d(A-A-T), * d(A-T-T), and RNA polymerase, the nearest neighbor results (Table 11) in the transcribed ribopolymer demonstrated dinucleotide sequences that were one-third each r(ApU) and r(UpA) and one-sixth each r(ApA) and r(UpU). When the highest molecular weight fraction (see Figure 2) of d(A-A-T), d(A-T-T), isolated from the large-scale reaction was used in a subsequent reaction for synthesis of more of the same polymer, the nearest neighbor results were incorrect such that, when [cx-~*P]~ATP was used as one of the substrates, 22 of the label was recovered in d-Ap and 78 in d-Tp. When [ C X - ~ ~ P I ~was T T Pused as one of the substrates, 56% of the label was found in d-Ap and 44x in d-Tp. However, when d(A-A-T),.d(A-T-T), with an average size of 129 base
z
pairs was used as primer, product was obtained corresponding in quantity to 44-fold synthesis and with nearest neighbor analyses correct for d(A-A-T),,.d(A-T-T), Figure 2 gives the results of fractionation according to mo-
-
0
TIME ( H O U R S 1
1 : Kinetics of polymer-directed synthesis of dA,. dT, (A), d(A-T),.d(A-T)n (B), and d(A-A-T),. d(A-T-T), (C). Each reaction mixture contained the following components in a total volume of 1 ml: 25 pmol of potassium phosphate buffer (pH 7.3, 5 pmol of MgC12, 1 pmol of 2-mercaptoethanol, 2.5 pmol each of dATP and dTTP, [ W ~ * P ] ~ TasT labeled P substrate, 50 mono-nmol of the polydeoxyribonucleotide template, and 150 units of E. coli DNA polymerase (phosphwellulosefraction). The incubation temperature was 15", and aliquots were analyzed as described under Experimental Procedure.
FIGURE
BIOCHEMISTRY, VOL.
12,
NO.
24, 1 9 7 3
5009
RATLIFF
TABLE 111: Physical
__
____
ff
a/.
Properties of the Three Isomeric Polymers.
Polymer
Tn, ("(2) (0.05 M Na+)
__
-.--
--
~
AA (nm) (Helix -+ Coil)
X (nm) (A,&,)
$2 0 -2 0 -0 5
258 262 260 5
--
__--
Slope, rn (Tn, = m In M 4-k )
___
____ - -
Buoyant - _Density In Cesium Chloride (g/ml)
In Cesium Sulfate (g/ml)
1 638 1 672 1 666 1 653"
1 319 1 427 I 420
-
dA, .dT, d(A-T),. d(A-T), d(A-A-T), . d(A-T-T), DNA (100% A.T)
61 54 56 58
8 5 7 6
.~
a
8 8 7 7 -
_
25 97 99 95 -
_
-
.
__
See Wells and Blair (1 967).
lecular size by chromatography through a Bio-Gel A-50m column using 0.05 M triethylammonium bicarbonate as eluent. The homopolymer dA,. dT, gave the largest molecular weight and sharpest size distribution followed. in turn, by d(A-T),. d(A-T), and d(A-A-T),. d(A-T-T), which came from the large reaction. Figure 3 gives representative melting curves for each of the polymers in 0.05 M Na+. The molarity dependence of the T , of d(A-A-T),, .d(A-T-T), was determined with high molecular weight polymer across a 100-fold range of Na+ concentrations at neutral pH. The linear least-squares fit to the data, expressed in degrees Celsius and moles per liter of Na+, is T,, = 7.99. (In M) 79.4. The slope (n? = 7.99) in this relationship is entered in Table 111 along with corresponding m values for dA, .dT, and d(A-T),. d(A-T), derived from the data of Wells et ul. (1970) and for DNA with 100% A . T pairs calculated from the equation of Frank-Kamenetskii (1971). The T,, values in 0.05 M Nai- for the three isomeric polymers as determined in Figure 3 are also entered in Table I11 along with the calculated T,,, in 0.05 M Na+ (Frank-Kamenetskii, 1971) for DNA with 100% A . T pairs. Also entered in Table I l l are
+
spectrophotometric data for the three isomeric polymers indicating X at A,,,, in the helix form and the change in X at A , , , as a result of melting the helix into the separated coils. The length dependence for the T,,,of d(A-A-T), . d(A-T-T),, was determined in 0.05 M Na+ across values of B ranging from 24 to 1000 base pairs. The linear least-squares fit to the data, expressedin degrees Kelvin and base pairs, is 10TT,,, = 3.021 2 .076/B . A scan of a buoyant density experiment in which E. coli DNA and d(A-A-T),. d(A-T-T), were centrifuged to equilibrium in neutral cesium chloride is shown in Figure 4. Using a value of 1.703 giml for the buoyant density of E. coli DNA (Vinograd et ul., 1963; Wells and Blair, 1967), the buoyant density of d(A-A-T),.d(A-T-T), was calculated to be 1.666 giml. Figure 4 also shows the buoyant densities of d(A-T),.d(A-T), and dA,.dT,. These polymers and marker E. coli DNA were run in separate cells in the four-hole An-F Ti rotor simultaneously with the cell which contained d(A-A-T),,.d(A-T-T),, and E. coli DNA. The values of 1.672 g/ml for d(A-T),.d
